The lateral heterogeneity of Glacial Isostatic Adjustment modelling across the Arctic

The Arctic has been key area for glacial isostatic adjustment (GIA) studies because it was covered by large ice sheets at the Last Glacial Maximum. Previous GIA studies applied mainly 1D Earth models. The few studies that did include 3D Earth structures have not considered the lateral heterogeneity differences across different regions of the Arctic. Here, using the latest standardized deglacial relative sea-level (RSL) databases from Norway and Russian Arctic, we investigate the effects of 3D structure on GIA predictions and explore the magnitudes of the lateral heterogeneity in both regions. The 3D Earth structure consists of 1D background viscosity model (ηo) and lateral viscosity variation, the latter is derived from the shear velocity anomaly from seismic tomography model and controlled by scaling factor (ß) denoting the magnitude of lateral heterogeneity. The Norway RSL database includes 413 sea-level index points (SLIPs), 175 marine limiting data and 433 terrestrial limiting data, while the Russian Arctic database includes 353 SLIPs, 78 marine limiting data and 92 terrestrial limiting data. We find 3D Earth structures have significant influences on RSL predictions and the optimal 3D model notably improves the fit with RSL data. However, we realize RSL data from Norway and Russian Arctic prefer different 3D structures to provide the best fits. The Russian Arctic database prefers a softer background viscosity model (ηo), but larger scaling factors (ß) than those preferred by Norway database. We further test the extent to which the 3D structure can be eliminated by refinement of ice model.